Note: Descriptions are shown in the official language in which they were submitted.
CA 02335331 2005-06-21
WO 99/66510 PCT/IL99/00324
METHOD AND SYSTEM FOR AUTHENTICATING
DIGITAL OPTICAL MEDIA
s FIELD AND BACKGROUND OF THE INVENTION
The present invention relates to a method and system for media
authentication and, more particularly, to a method and system for
authenticating
digital optical media.
Digital optical media includes, but is not limited to, media such as
to Compact Disc (CD), Compact Disc Read-Only Memory (CD-ROM), and Digital
Video Disc (DVD). Digital optical media are well-known in the art and have
become the media of choice for a broad variety of important data storage and
information distribution applications. In particular, certain proprietary
materials
such as computer software, specialized data, and audio/video content are
~s commonly sold and distributed on digital optical media.
Digital optical media technology is established according to a series of
international publications, herein referred to as "standards'e
For example, some
common standards applicable to CD's include: the International Standards
2o Organization ((SO) standard 9660 entitled "Information Processing -- Volume
and File Structure of CD-ROM for information interchange, ISO Standard
13490-1 ", the international Electrotechnique Commission (CEI-IEC) standard
908, generally conforming to what is known as the "Red Book", and ISO/IEC
10140, generally conforming to what is known as the "Yellow Book".
2s Software and document data may be read and utilized by a computer
from digital optical media, and there are widely-available players for reading
data from digital optical media and using this data to reconstruct audio,
visual,
text, and audio-visual information. The term "player" herein denotes any
device
which is able to read data from digital optical media. Players include, but
are not
30 limited to, CD players, CD-ROM multi-media players, game-playing systems,
CA 02335331 2000-12-15
WO 99/66510 PCT/IL99/003~4
and DVD players, which can reproduce sound, images, and text from data
stored on digital optical media. Some computers are also configured with
hardware and software capable of accessing digital optical media, to duplicate
the functionality of CD players, CD-ROM multi-media players, game-playing
s systems, and DVD players.
Unfortunately, it is easy to copy proprietary material from an original
digital optical medium and thereby produce an unauthorized copy whose sale
and distribution cannot be controlled by the owner of the proprietary
material.
Individual users can freely copy such proprietary material using low-cost
ro consumer devices such as CD-R recorders, and it is also possible to
mass-produce unauthorized copies of proprietary material using commercial
mastering equipment. It is usually difficult or impossible to enforce
copyright
laws in such cases, and the legitimate owner of the proprietary material is
thereby deprived of the legal right to control the sale and distribution of
the
is proprietary material. The term "original" as used herein refers to an
instance of a
digital optical medium which has been authorized by and issued under the
control of the owner of the proprietary material recorded thereon. In
contrast,
the term "unauthorized copy" herein denotes an instance of a digital optical
medium which has derived from an original digital optical medium via copying
2o that has neither been authorized by nor is under control of the owner of
the
proprietary material recorded thereon.
Thus, there is a widely-recognized need for a means of distinguishing an
original digital optical medium from an unauthorized copy, and especially,
there
is a widely-recognized need for a means of automatically distinguishing an
2s original digital optical medium from an unauthorized copy. The term
"automatically distinguish" herein denotes a means of distinguishing an
original
digital optical medium from an unauthorized copy in such a way that does not
2
CA 02335331 2000-12-15
WO 99/66510 PCT/IL99/00324
require visual inspection or other human intervention. In particular,
automatically
distinguishing an original digital optical medium from an unauthorized copy
should be feasible by a player as it reads the digital optical medium. That
is,
whatever distinguishing features are placed on an original digital optical
medium
s for identification as original, the distinguishing features must be
machine-readable. The terms "authenticate" and "authenticating" herein refer
to
any process by which an undetermined digital optical medium corresponding to
an original digital optical medium can be differentiated to be an original
digital
optical medium as distinct from an unauthorized copy of a original digital
optical
to medium. The term "undetermined" herein denotes that a specific instance of
a
digital optical medium is not yet known to be an original digital optical
medium
as distinct from an unauthori~_ed copy. The term "corresponding to" herein
denotes that a specific instance of a digital optical medium contains the same
functional data as a given original digital optical medium.
is In addition to allowing a player to identify the digital optical medium
being
played as an original rather than an unauthorized copy, a method for
authenticating an undetermined digital optical medium can allow the player to
selectively access proprietary material only if the digital optical medium is
an
original, and deny access to the' proprietary material if the digital optical
medium
2o is an unauthorized copy. Such selective access is a means of copy
protection,
and can be implemented through various techniques known in the art, such as
by encrypting the proprietary material and storing a decryption key on the
original digital optical medium in such a way that the decryption key is not
readily copyable. The presence of a valid decryption key on a digits! optical
2s medium therefore is intended to automatically distinguish the digital
optical
medium as original, and moreover to provide copy protection by allowing the
player access to the proprietary material only if the digital optical medium
is
3
CA 02335331 2000-12-15
WO 99/66510 PCT/IL99/00324
original. The term "copy protection" herein denotes any method or system which
either prevents the making of an unauthorized copy or renders an unauthorized
copy useless for its intended application.
There are currently a nurnber of schemes which attempt to automatically
s distinguish an original digital optical medium from an unauthorized copy by
storing a special pattern, such as a decryption key, on the digital optical
medium
in such a way that the special pattern is not readily copyable. For example,
US
Patent number 5,400,319 to Fite et al. ("CD-ROM with Machine-Readable I.D.
Code") discloses the use of a laser to selectively destroy portions of the
to reflective layer of the CD-ROM, thereby creating addressable defects in
which a
serial number may be encodE:d. US Patent number 5,563,947 to Kikinis
("CD-PROM") discloses a similar use of a laser to physically damage selected
sectors and thereby produce a pattern of unreadable sectors in which a
decryption key may be stored. And US Patent number 5,703,858 to Mitchell et
~s al. ("System for Encoding a Glass Master to Enable Detection of a
Counterfeit
Optical CD-ROM") discloses the use of a high-frequency random modulation of
a laser to produce random defects in the CD-ROM at the master level by
selective destruction of predetermined sectors. All of these techniques, as
well
as other currently-available commercial techniques for copy-protecting digital
20 optical media, involve creating damaged, invalid, or otherwise unreadable
portions of the digital optical medium. The theory behind such techniques is
that
ordinary consumer recording equipment is not intended to be able to produce
defects in the copies and therE~fore an unauthorized copy produced on such
equipment should lack the unreadable areas in which the decryption key or
2s other pattern is encoded. If this were in fact reliably the case, then such
techniques would provide means far authenticating an undetermined digital
optical medium. The present applicants, however, have found that it is
possible
4
CA 02335331 2000-12-15
WO 99/66510 PCT/IL99/003~4
to reproduce ~ unreadable sectors in a copy by using ordinary consumer
recording equipment with the appropriate software. Thus, a person with access
to the proper software would be able to easily defeat such schemes relying on
unreadable sectors and thereby create an unauthorized copy of a digital
optical
s medium which will be incorrectlly discerned by these authenticating schemes
to
be an original digital optical medium. In effect, all the schemes currently
known
in the art for providing copy protection of digital optical media by encoding
patterns in damaged or unreadable data areas are of limited value in
distinguishing an original digital optical medium from an unauthorized copy
and
to offer only limited protection against making unauthorized copies.
Therefore, it would be highly advantageous to have a method and system
for authenticating an undeterrnined digital optical medium which cannot be
defeated utilizing commercially-available copying equipment, regardless of the
software employed. This goal is met by the present invention.
is
SUMMARY OF THE INVENTION
The present invention relies on the fact that the standards specify certain
20 error-correction limits on the abiility of digital optical media to
tolerate errors. The
standards provide for error correction by specifying error-correction data
subunits which contain redundant information that can be used to reconstruct
damaged or missing data. They term "data subunit" herein denotes any set of
data symbols grouped together according to the appropriate standards for the
2s digital optical medium. The term "data symbol" herein denotes the primitive
data
element defined according to the appropriate standards for the digital optical
medium. For example, the standards for Compact Disc specify data symbols
s
CA 02335331 2000-12-15
WO 99/66510 PCT/1L99/00324
corresponding to a byte (8 binary digits). The term "error-correction data
subunit" herein denotes any set of data symbols grouped together far
error-correction purposes according to the appropriate standards for the
digital
optical medium. Typically, an error-correction data subunit may be contained
s within a larger data subunit, arrd all the data symbols and data subunits of
a
digital optical medium are addressable. The term "addressable" herein refers
to
the property of a data symbol or data subunit on a digital optical medium that
allows a player of the digital optical medium to locate and read a specific
predetermined data symbol or data subunit. That is, the specific predetermined
~o data symbol or data subunit has a unique machine-accessible location or
"address". For example, a data subunit of a Compact Disc is a "sector", an
error-correction data subunit of .a Compact Disc is a "C2 codeword", and each
such data subunit has a unique numerical address. At or below the
error-correction limit for an error-correction data subunit, the player must
be able
is to correct errors and read the error-correction data subunit error-free.
Above this
error-correction limit, however, the error-correction data subunit (or a
containing
data subunit) with errors cannot be read, or optionally (depending on the
player)
the error-correction data subunit or the containing data subunit will be read
with
reported or detectable errors. F=or example, the standards for Compact Disc
2o furthermore specify that a C2 codeword be correctable with up to 2 errors.
That
is, the error-correction limit for a C2 codeword is 2. At or below this
error-correction limit, the player can correct errors to read a C2 codeword
without any errors at all. Above 'this error-correction limit, however, the
player is
unable to correct the errors and will report an "E32" uncorrectable error
2s condition. This error may affect the readability of the entire sector in
which the
C2 codeword is located.
6
CA 02335331 2005-06-21
WO 99/66510 PCT/IL99/00324
Figure 3, to which reference is now briefly made, shows a prior art digital
optical medium having a data subunit 110 containing a number of
error-correction data subunits 120, 130, 140, and 150. Figure 4, to which
reference is now briefly made, shows prior art error-correction data subunit
150
s containing a number of data symbols 200, 210, 220, and 230.
The term "bistable" herein denotes data recorded on digital optical media
which randomly assumes one of at least two distinctly different states upon
reading. Thus, for example, a bistable data symbol is a data symbol which
randomly assumes one of at least two distinctly different values upon reading.
~o In contrast, the term "monostable" herein denotes data recorded on digital
optical media which always assumes the same state upon reading. Thus, a
monostable data subunit always reads with the same data contents no matter
how many times it is read. By recording special bistable data symbols and
other
deliberate errors within error-correction data subunits, it is possible to
record
is bistable error-correction data subunits which upon reading randomly appear
either to be at or to be above the error-correction limit for the error-
correction
data subunit. The term "stability" herein denotes the property of any specific
data subunit as being either monostable or bistable. Ordinarily, equipment
used
to record digital optical media is capable of writing only monostable data.
2o Special hardware equipment is required to write bistable data on digital
optical
media.
A bistabie data symbol is, in effect, a data symbol with an invalid code that
produces an ambiguous output. Methods of creating such a bistable data
symbol are covered in corresponding International Patent Application
Publication
2s No. W098/08180 by Sollish et al. and published February 26, 1998.
CA 02335331 2000-12-15
WO 99/66510 PCT/IL99/00324
When writing a bistabie error-correction data subunit containing a bistable
data symbol, one of the values of the bistable data symbol is the correct
value
that the data symbol should take, and when the bistable data symbol assumes
this correct value the bistable error-correction data subunit will not exceed
the
s error-correction limit and will therefore be readable without error. The
other
values of the bistable data symbol, however, are erroneous values, and when
the bistable data symbol .assumes an erroneous value, the bistable
error-correction data subunit containing the bistable data symbol will exceed
the
error-correction limit, and will therefore be either unreadable or will be
read with
to reported or detectable errors. If necessary, additional bistable data
symbols
and/or erroneous monostable data symbols are written to the data subunit in
order to either exceed or not exceed the error-correction limit. Since writing
bistable data symbols requires special hardware equipment, ordinary consumer
recorders of digital optical mE;dia cannot reproduce them, regardless of the
is software employed. Thus, an unauthorized copy of a digital optical medium
will
contain only ordinary monostable data symbols, which always yield the same
value upon reading. Consequently, an unauthorized copy of a digital optical
medium will contain only ordinary monostable data subunits. An unauthorized
copy can therefore be automatically distinguished from an original digital
optical
2o medium, since the original digii:al optical medium has at least one
bistable data
subunit which randomly reads either without errors or with errors, whereas
every
data subunit of an unauthorized copy is monostable and will either always read
the same way ---- either always without errors or always with errors.
By recording an original digital optical medium with a bistable data subunit,
2s it is possible to authenticate an undetermined digital optical medium
corresponding to the original digital optical medium by testing the
undetermined
digital optical medium to ascertain the stability of the data subunit with the
same
s
CA 02335331 2000-12-15
WO 99/66510 PCT/IL99/00324
address as a bistable data subunit on the original digital optical medium. If
the
data subunit on the undetermined digital optical medium is bistable then the
undetermined digital optical medium is an original digital optical medium.
Otherwise, if the data subunit is monostable, then the undetermined digital
s optical medium is an unauthorized copy.
Therefore, according to the present invention there is provided a method
for authenticating an undetermined digital optical medium corresponding to an
original digital optical medium, including the steps of: (a) writing at feast
one
addressable bistable data subunit on the original digital optical medium, the
~o addressable bistable data subunit having a predetermined address; (b)
testing
the stability of an addressable data subunit of the undetermined digital
optical
medium having the same address as the predetermined address; and ('c)
determining the undetermined digital optical medium to be an original digital
optical medium if the result of the testing is that the stability is bistable,
and
is determining the undetermined diigital optical medium to be an unauthorized
copy
if the result of the testing is that the stability is monostable.
According to a preferred embodiment of the present invention, it is
possible to determine the stabiliioy of a data subunit by making multiple
readings
of the data subunit. If the multiple readings have differing results (some
with
2o errors and some without error;s), then the data subunit is bistable and the
undetermined digital optical medium being authenticated is an original digital
optical medium. Otherwise, if tine multiple readings have identical results
(all
with errors or all without errors), then the data subunit is monostable and
the
undetermined digital optical mE:dium being authenticated is an unauthorized
2s copy. In another embodiment of the present invention, the data returned
from a
series of read attempts is connpared. For certain players, a bistable data
subunit will return different data values on subsequent read operations, and
this
9
CA 02335331 2000-12-15
WO 99/66510 PCT/IL99/00324
fact can sometimes be used to determine the stability. In yet another
embodiment of the present invention, a player's built-in retry capabilities
are
utilized to attempt to read the .data subunit. Some players have the ability
to
make repeated attempts to head a data subunit if the initial attempt is
s unsuccessful on account of errors. The time required to read a data subunit
corresponding to a bistable data subunit on an original digital optical medium
and return valid data is measured and compared against that required to read a
known monostable data subunit. If the time to read the data subunit
corresponding to a bistable dai:a subunit is not substantially greater than
the
Io time to read a known monostable data subunit, the data subunit
corresponding
to a bistable data subunit on an original digital optical medium is a
monostable
data subunit and the digital optical medium is determined to be an
unauthorized
copy. Likewise, if the player cannot read the data subunit at all, then it is
also a
monostable data subunit and the digital optical medium is determined to be an
is unauthorized copy. Only if the player takes substantially longer to read
the data
subunit corresponding to a bistable data subunit on an original digital
optical
medium than for reading a knovvn monostable data subunit is the data subunit
bistable, in which case the digital optical medium is determined to be an
original.
to
CA 02335331 2000-12-15
WO 99/66510 PCT/IL99/00324
BRIEF DESCRIPTION OF THE DRAWINGS
The invention is herein described, by way of example only, with reference
to the accompanying drawings, wherein:
s Figure 1 shows the basic elements of authenticating digital optical media.
Figure 2 is a flowchart showing the steps in authenticating an
undetermined digital optical medium according to the present invention.
Figure 3 shows the makeup of a data subunit on a prior art digital optical
medium.
~o Figure 4 shows the makeup of an error-correction data subunit of a prior
art digital optical medium.
Figure 5 illustrates a general example of the makeup of a bistable
error-correction data subunit of a digital optical medium according to the
present
invention.
is Figure 6 is a flowchart showing the steps in writing a bistable data
subunit
according to the present invention.
Figure 7 is a flowchart showing the steps in determining data subunit
stability by multiple reading operations to check for errors.
Figure 8 is a flowchart showing the steps in determining data subunit
2o stability by multiple reading operations to check for data.
Figure 9 is a flowchart showing the steps in determining data subunit
stability by timing measurements.
Figure 10 is a flowchart showing the steps in an alternative determination
of data subunit stability by timing measurements.
n
CA 02335331 2000-12-15
WO 99/66510 PCTIIL99/00324
DESCRIPTION OF THE PREFERRED EMBODIMENTS
The present invention is of a method for automatically distinguishing an
original digital optical medium from an unauthorized copy. Specifically, the
s present invention can be used to authenticate an original digital optical
medium,
either purely for identification c~f unauthorized copies, or to provide a
basis for
copy-protecting proprietary material on an original digital optical medium.
The principles and operation of a method and system according to the
present invention may be better understood with reference to the drawings and
to the accompanying description.
Referring now to the drawings, Figure 1 shows what happens an original
digital optical medium 10 containing a bistable data subunit 12 with an
address
30 undergoes a copy operation 15 to produce an unauthorized copy 20 of a
digital optical medium corresponding to original digital optical medium 10. It
is
~s seen that unauthorized copy i!0 contains a monostable data subunit 22 also
having address 30. The only difference between the original digital optical
medium 10 and unauthorized copy 20 that can be automatically distinguished is
that data subunit 22 in the unauthorized copy is monostable rather than
bistable
as is data subunit 12 in the original digital optical medium.
2o Figure 2 is a flowchart showing the general steps of the authentication
method according to the present invention. In a step 50, a bistable data
subunit
is recorded onto the original digital optical medium at an address A. In a
step
60, the undetermined digital optical medium is tested to determine the
stability
of the data subunit at address A. Then, at a decision point 70, the method
2s concludes with a result 80 that the undetermined digital optical medium is
an
original digital optical medium if the stability is bistable and with a result
90 that
12
CA 02335331 2000-12-15
WO 99/66510 PCT/IL99/003Z4
the undetermined digital optical medium is an unauthorized copy if the
stability
is not bistable (i.e., the stability is monostable).
Figure 5 illustrates a general example of the makeup of a bistable
error-correction data subunit 250 according to the present invention. Bistable
s error-correction data subunit 250 has an error-correction limit of N,
meaning that
error-correction data subunit 250 may contain up to N errors in arbitrary data
symbols and still be readable without error. That is, up to N arbitrary data
symbol errors may be corrected within error-correction data subunit 250. In
this
example, bistable error-correction data subunit 250 contains data symbols 252,
Io 254, 256, 258, 262, 264, 266, 268, 272, 274, 276, and 278. The ellipsis
(...)
indicates that additional data symbols may in general also be present. In this
example, the bistable properties of bistable error-correction data subunit 250
are
based on the fact that data symbols 256 and 278 are bistable data symbols.
Data symbols 256 and 278 were arbitrarily chosen from the data symbols in
is error-correction data subunit 2.50 to be written as bistable, and are
contained
within a subgroup 284. The ellipsis (...) within subgroup 284 indicates that
in
general additional bistable data symbols may be contained there. Moreover,
data symbols 252, 264, and 2T4 were arbitrarily chosen from the data symbols
in error-correction data subunii: 250 to be written as erroneous data symbols,
zo and are contained within a subgroup 282. The ellipsis (...) within subgroup
282
indicates that in general additional erroneous data symbols may be contained
there. Subgroup 282 and subgroup 284 constitute a group 280 which contains
more than N data symbols (the number of data symbols in group 280 is > N). In
general, subgroup 282 is the complement of subgroup 284 in group 280. That
2s is, group 280 consists exactly of the union of subgroup 282 and subgroup
284.
Likewise, subgroup 284 is thE: complement of subgroup 282 in group 280.
Furthermore, subgroups 282 and 284 are mutually exclusive. That is, no data
I3
CA 02335331 2000-12-15
WO 99/66510 PCT/IL99/00324
symbol in subgroup 282 is also contained in subgroup 284. Moreover, there
must be at least one bistable data symbol in subgroup 284, and the number of
erroneous data symbols in subgroup 282 cannot be greater than N. Subgroup
282, however, may be empty, in which case group 280 consists entirely and
s solely of subgroup 284 (that is, group 280 equals subgroup 284). In
practice, a
variety of arbitrary selections of data symbols for subgroups 282 and 284 is
possible. These conditions insure that error-correction data subunit 250 will
be
bistable. Depending on the read states of bistable data symbols in subgroup
284, error-correction data subunit 250 reads with N or fewer errors, in which
to case the read operation returns error-correction data subunit 250 without
errors,
or error-correction data subunit 250 reads with more than N errors, in which
case the read operation is unsuccessful because the error-correction limit of
error-correction data subunit 250 is exceeded.
Figure 6 is a flowchart showing the method according to the present
is invention for writing a bistable data subunit such as bistable error-
correction
data subunit 250 (Figure 5). In a step 290 a number > N of data symbols are
selected for group 280 (Figure ;i). Then, in a step 292 at least one data
symbol
of group 280 (Figure 5) is selected for bistable subgroup 284 (Figure 5).
Then,
in a step 294 the data symbols of bistable subgroup 284 (Figure 5) are
replaced
2o with bistable data symbols. Then, in a step 296 the complement of bistable
subgroup 284 (Figure 5) is selected for erroneous subgroup 282 (Figure 5).
Finally, in a step 298 the data symbols of erroneous subgroup 282 are replaced
with erroneous data symbols.
Figure 7 is a flowchart showing the steps of an embodiment according to
zs the present invention for determining the stability of a data subunit of
the data
subunit at address A. Recalling that a bistable data subunit randomly returns
different data when read, it is seen that it is necessary to classify the
returned
14
CA 02335331 2000-12-15
WO 99/66510 PCT/IL99/00324
data into groups of different potential responses. If the actual responses of
the
read operations result in returned data of different classifications, then the
data
subunit is determined to be bistable. Otherwise, if all the actual responses
of the
read operations result in returned data of the same classification, then the
data
s subunit is determined to be monostable. For this example, the different
potential
responses are classified as being either "error-free" or "error". The steps of
this
portion of the method according to the present invention therefore require two
counters, an error counter 300 and an error-free counter 310. In a first step
320,
error-free counter 310 and error counter 300 are zeroed. Following this is a
loop
to 330-380 of N read operations.. The purpose of the N read operations is to
determine the stability by repeatedly attempting to read the data subunit at
address A. If the data subunit is bistable, the number N must clearly be
greater
than 1, for otherwise it will be irnpossible to detect different values upon
reading
the data subunit. The larger N becomes, the more likely will be the
probability
is that a bistable data subunit will return different values, and the more
statistically
significant will be a result that: the data subunit is monostable. On the
other
hand, the larger N becomes, the more time is required for reading. In
practice, a
value of N= 5 is usually sufficient. Within loop 330-380 there is a read
operation
340 which attempts to read i:he data subunit at address A. Following is a
2a decision point 350 where the results of the read attempt are analyzed. If
the
data subunit is readable without error, a step 360 increments error-free
counter
310. Otherwise, if the data subunit is not readable without error, a step 370
increments error counter 300. After loop 330-380 completes, a decision point
390 checks error-free counter 310 for a zero count. If error-free counter 310
is
2s zero, then all of the read responses were the same {all erroneous), and the
result of the test is that the data subunit with address A is monostable, at a
result 420. Otherwise, a decision point 410 checks error counter 300 for a
zero
is
CA 02335331 2000-12-15
WO 99/66510 PCT/IL99/00324
count. If error counter 300 is zero, then all the read responses were the same
(all error-free} and the result of the test is that the data subunit with
address A is
monostable, also at result 420. Otherwise, some of the read responses were
erroneous and some were error-free, and the result of the test is that the
data
s subunit with address A is bistable, at a result 430. Only if both error
counter 300
and error-free counter 310 are non-zero is the data subunit determined to be
bistable.
Figure 8 is a flowchart showing the steps of another embodiment
according to the present invention for determining the stability of a data
subunit
to of the data subunit at address A. This embodiment also utilizes multiple
read
operations, but instead of classifying the outcome of the individual read
operations as "error-free" or "error", the values of the read data are used.
The
values of the read data for the different read operations are compared, and if
there are any differences, then the data subunit is determined to be bistable.
is The present applicants have determined that certain players return an
indication
of "effor-free" reading, but nevE~rtheless fail to perfectly correct the
erroneous
data of the bistable data subuniit. Therefore, it is possible with these
players to
detect a bistable data subunit by comparing the data read. Otherwise, if all
the
read operations result in identic~ai data, the test is inconclusive. In this
example,
2o the test is performed by having a first read buffer 500 and a subsequent
read
buffer 510. The data from reading the data subunit at address A is loaded into
first read buffer 500 in a read operation 520. Within loop 530-580 there is a
read
operation 540 which reads the data subunit at address A and loads the data
into
subsequent read buffer 510. Following is a decision point 550 where the data
2s stored in first read buffer 500 is compared with the data stored in
subsequent
read buffer 510. If the data in first read buffer 500 differs from that of the
data in
subsequent read buffer 510, then the test determines that the data subunit is
16
CA 02335331 2000-12-15
WO 99/66510 PCT/IL99/00~24
bistable in result 630. At the end of loop 530-580, the test is deemed
inconclusive in a result 620.
Figure 9 is a flowchart showing the steps of yet another embodiment
according to the present invention for determining the stability of a data
subunit
s of the data subunit at address A. This embodiment utilizes timing
measurements for a single read operation. There are two registers for storing
time measurements, a monostable time register 700 and an undetermined time
register 710. At a time measurement step 720, a measurement is made of the
elapsed time required to read a known monostable data subunit, and this time
to measurement is placed in monostable time register 700. Then, at a time
measurement step 730, a measurement is made of the elapsed time required to
read the data subunit at address A, and this time measurement is placed in
undetermined time register 710. Next, at a decision point 740 if the read
operation in time measurement step 730 is not a successful read response, the
1:~ test terminates as inconclusivE~ at a result 750. Otherwise, at a decision
point
760, the contents of undetermiined time register 710 are compared against that
of monostable time register 700. If the contents of undetermined time register
710 are substantially greater than the contents of monostable time register
700,
then the test determines tha~~l the data subunit is bistable in a result 770.
2o Otherwise, if the contents of undetermined time register 710 are not
substantially greater than the contents of monostable time register 700, then
the
test determines that the data s~ubunit is monostable in a result 780. A first
time
measurement T~ is herein defined to be "substantially greater than" a second
time measurement z2 if the ratio ~' is of the order of at least 1.5.
Tz
2:. Figure 10 is a flowchart showing the steps of an embodiment according to
the present invention, similar to that shown in Figure 9, except that at
decision
point 740, if the read operation in time measurement step 730 is not a
t~
CA 02335331 2000-12-15
WO 99/66510 PCT/IL99/00324
successful read response, then the test determines that the data subunit is
monostable in result 780. In all other respects, the steps are the same as
shown
in Figure 9.
Combinations of the tests described above can also be used. F'or
s example, a data read test (Figure 8) can be employed, and if this test is
inconclusive, an error test (Figure7) can be then invoked to resolve the
issue.
While the invention has been described with respect to a limited number of
embodiments, it will be appreciated that many variations, modifications and
other applications of the invention may be made.
rs
